- From: James M Snell <jasnell@gmail.com>
- Date: Mon, 11 Jun 2012 17:21:52 -0700
- To: ietf-http-wg@w3.org
My apologies... accidentally responded directly to Mark instead of the group... Another possible approach to early implementation is to designate a range of experimental, non-production status codes for early development purposes that MUST NOT be used in production... once the draft progresses to a reasonable stage (well beyond -01, the real status can be assigned to the spec by the registrar rather than by the spec author. I know schemes like this can tend to be problematic (e.g. all those damn X- HTTP headers) so I'm not sure if it's a path we should go down, but it's an idea at least. ---------- Forwarded message ---------- From: Mark Nottingham <mnot@mnot.net> Date: Mon, Jun 11, 2012 at 4:58 PM Subject: Re: Proposing Status Codes To: James M Snell <jasnell@gmail.com> On 12/06/2012, at 9:55 AM, James M Snell wrote: > Perhaps one solution is to require the registrar to assign the status code upon completion... However, with any such late assignment approach, we run the risk of blocking important, valuable early implementation. Yeah, I thought about that, but until the registry is actually reflecting deployment practice, I think there needs to be more oversight. What I'd like to see is having a -00 draft with 5xx (for example) specified, but maybe -01 (or -02, or..) going ahead and picking one, based on community input, once it's clear that there's real interest. -- Mark Nottingham http://www.mnot.net/
Received on Tuesday, 12 June 2012 00:22:42 UTC